305 results about "Noise floor" patented technology

The invention relates to the technical field of artificial intelligence, and provides a risk content identification method based on a multi-modal large model. The method comprises the following steps: identifying a forged part in audio; extracting bottom noise features and high-frequency features of the target face image, inputting the features into the image segmentation model, and positioning a forged area; dicing the target face image, and inputting a local area image and a global image obtained by dicing into a visual encoder to extract visual features; calculating the attention of the text features and the global image features on the local image features, and discarding the local image features with small attention; and inputting the output of the audio segmentation model and the image segmentation model, image features and questions into a large language model, and summarizing risk points. According to the method, multiple data sources are integrated, risk identification is carried out based on the multi-modal large model, the accuracy and robustness of identification are improved, multiple fraud means can be effectively coped with, and the problems that in the prior art, data of multiple modals cannot be coped with and interpretability is lacked are solved.

The invention discloses a signal identification method based on broadband spectrum detection, and the method comprises the steps: carrying out the power spectrum estimation and signal detection of broadband IQ data through an FPGA, obtaining a signal frequency point parameter, and configuring a DDC module to complete digital down conversion; performing power spectrum density estimation and noise floor estimation on the narrowband IQ signal on the upper computer; preliminary classification is carried out by calculating a ratio parameter C1 of a maximum value and a second maximum value of a frequency spectrum, and when C1 is greater than a threshold, AM and a single carrier signal are discriminated by combining a frequency spectrum characteristic parameter C2 and an amplitude spectrum characteristic parameter C3; when C1 is smaller than or equal to a threshold, signal detection is carried out by adopting a double-threshold algorithm, and merging and screening in a multi-signal scene are completed; classifying based on a normalized zero center instantaneous amplitude spectrum density maximum value parameter C4, judging FM, MSK and FSK signals when the C4 is less than a threshold, and judging BPSK, QPSK, 8PSK and 16QAM signals when the C4 is greater than or equal to the threshold; according to the method, a multi-parameter joint judgment strategy is adopted, and the accuracy and the real-time performance of signal recognition are remarkably improved.

The invention discloses a periodogram and incoming wave direction finding fused multi-unmanned aerial vehicle signal identification method and system, and relates to a signal processing technology, and the method comprises the steps: obtaining to-be-identified signals with time domain aliasing and frequency domain non-aliasing; suppressing the noise of the to-be-identified signal by using an average periodogram method and smooth filtering, and calculating a bottom noise value; according to the calculated bottom noise value, setting a judgment threshold to detect a target unmanned aerial vehicle signal; estimating the start-stop frequency and bandwidth of the detected target unmanned aerial vehicle signal; performing signal separation by using an FIR filter according to the estimated starting and ending frequency and bandwidth; and executing incoming wave lateral direction based on the separated target unmanned aerial vehicle signal so as to complete signal identification. According to the method, unmanned aerial vehicle signal detection is completed through an average periodogram method, unmanned aerial vehicle weak signals which are sparsely distributed in space are further obtained through a direction finding method, and weak signal unmanned aerial vehicles which are far in distance are detected.

A reference system is provided for generating a clock signal with a tunable noise pedestal for driving a signal generator. The reference system includes a reference source configured to generate a reference signal and a multiplier chain, including a noise pedestal generator. The noise pedestal generator includes a noise pedestal attenuator configured to attenuate the reference signal to degrade a noise floor of the reference signal, a VGA configured to adjust the attenuated reference signal to a desired signal level, a bandpass filter configured to filter out excess noise from the adjusted reference signal to form a noise pedestal, and a switch arrangement configured to selectively input the reference signal to a first path including the noise pedestal attenuator and the VGA, and a second path bypassing the noise pedestal attenuator and the VGA. The clock signal includes the noise pedestal when the first path is selected.

The invention relates to an adaptive noise floor energy detection method and system based on FFT frequency spectrum accumulation, and the method comprises the steps: carrying out the FFT frequency spectrum accumulation of an electromagnetic signal sampled by an ADC in an FPGA, and carrying out the channel division of a broadband frequency spectrum into sub-bands; counting and estimating the top and the bottom of the noise by sub-bands one by one; carrying out median filtering processing, subtraction and modular operation according to the obtained data to obtain optimal values of the energy bottom and the noise energy fluctuation range of each channel, carrying out summation to obtain a noise floor of each sub-band, and carrying out median filtering once to obtain a final noise floor value of each sub-band; and a noise factor is introduced to obtain a final adaptive constant false alarm threshold, and adaptive detection of noise floor energy is realized through the threshold. According to the invention, a parallel operation architecture based on the FPGA is adopted, so that the method is simple in calculation and high in sensitivity, and has the capability of detecting each sub-band signal at the same time.

The invention provides an ultrasonic sensor temperature confirmation method. The ultrasonic sensor temperature confirmation method comprises the steps that the body temperature T2 of an ultrasonic sensor installed on an automobile is collected; the acquisition result is analyzed, if the body temperature T2 is successfully acquired, the transducer temperature T3 is acquired, and the final temperature of the ultrasonic sensor is calculated according to the transducer temperature T3 and the body temperature T2; and if the body temperature T2 is not successfully collected, the outdoor temperature T1 of the automobile is obtained, and the final temperature of the ultrasonic sensor is calculated according to the transducer temperature T3 and the outdoor temperature T1 of the automobile. In consideration of different key parameters, such as transducer resonant frequency, impedance, sensor gain, ground noise and the like, better performance can be embodied after temperature influence compensation, so that measurement of the temperature of the ultrasonic sensor has an important effect on the stability of the whole vehicle body.

The invention discloses a fault detection device and method and a photovoltaic power generation system, and belongs to the technical field of electric power. The fault detection device comprises a sampling module used for sampling an output signal of an output cable of a solar cell panel to obtain a sampling signal; the output signal comprises a current bottom noise ripple signal and a fault ripple signal; the signal processing module is used for carrying out wave trapping and conditioning processing on the sampling signals, filtering out signals of a target frequency band in the sampling signals and amplifying signals of other frequency bands to obtain characteristic signals; the target frequency band is a frequency band where the characteristic noise signal coincides with a noise signal generated by the target fault; and the control module is used for detecting whether a fault exists based on the characteristic signal, the background noise model and the fault signal model. According to the fault detection device and method and the photovoltaic power generation system disclosed by the invention, the sampling notch and conditioning processing are combined, the switching frequency noise, the frequency multiplication noise, the current total harmonic distortion noise and the like are suppressed, and the detection accuracy can be improved.

The invention discloses a multi-bandwidth spectrum data signal adaptive extraction method based on a signal-to-noise ratio threshold, and relates to the technical field of spectrum signal extraction. The frequency spectrum data collection covers 30MHz-6GHz, the sampling rate is 20MHz, 16-bit quantization is realized, and the hardware fluctuation is eliminated through the dual-channel design; according to noise floor estimation, 50 frames of no-signal samples are taken, mean value calculation is carried out through a sliding window, then median filtering is carried out, and the noise power and confidence interval of each frequency point are obtained; basic values are set for a signal-to-noise ratio threshold according to frequency bands, dynamic adjustment is performed in combination with real-time noise standard deviation, and compensation is added to a known frequency band; for multi-bandwidth detection, wavelet packet decomposition is adapted to 100kHz-5MHz, candidate signals are marked when energy detection exceeds a threshold, and a range is determined by morphological filtering; extracting parameters such as center frequency, analyzing time-frequency characteristics through short-time Fourier transform, and calculating 4-order cumulants to establish feature vectors; coherent accumulation times are adjusted according to a signal-to-noise ratio, energy is superposed through phase alignment, and a purified signal is output through band-pass filtering. According to the invention, the signal detection precision and adaptability in a complex environment are improved, and misjudgment is reduced.

The invention discloses a self-adaptive large dynamic detection noise floor suppression system and a control method thereof. The system comprises a filtering selection front-end module, a detection synthesis module, a frequency measurement receiving module and an FPGA feedback control module. The method is composed of four main parts, namely a filtering selection front end, detection synthesis, frequency measurement receiving and FPGA feedback control. Wherein the filtering selection front end and the frequency measurement receiving end receive two groups of external radio frequency signals at the same time. And frequency measurement is carried out through a mature single-bit receiving scheme in frequency measurement receiving, and a result is sent to the feedback control module, so that switching of a filtering channel and suppression of a noise floor are realized. According to the invention, the broadband large dynamic DLVA detection sensitivity is obviously improved.

The invention discloses a high-sensitivity low-noise hydrophone system structure based on a four-core optical fiber. The high-sensitivity low-noise hydrophone system structure comprises a laser, an optical modulation and beam splitting module, a four-core optical fiber hydrophone and a phase demodulation and noise suppression processing module. The four-core optical fiber hydrophone comprises a first reference optical fiber unit, a first cascade sensitization channel, a second reference optical fiber unit and a second cascade sensitization channel, the first reference unit and the first cascade sensitization channel output interference optical signals I1 and I2 respectively, and the second reference unit and the second cascade sensitization channel output interference optical signals I3 and I4 respectively. The device is applied to the field of hydrophones, a four-core optical fiber is used for constructing an ultra-long convolution sensitization optical path, then inter-channel noise correlation is used for constructing a multi-channel interference complementation optical path, sensitivity improvement and self-suppression of common-mode noise are achieved at the same time through combined processing, an additional optical fiber interferometer does not need to be introduced, the noise is irrelevant to a transmission optical path, and the cost is low. And the complexity and the background noise of the system are reduced.

The invention provides a data processing and output circuit and method for a high-energy particle detection pixel sensor chip, and relates to the technical field of high-energy particle detection, and the method comprises the steps that each data processing unit receives ADC output of multiple channels, and carries out the validity discrimination of the received original data of each channel in a polling mode; effective case information in each data processing unit is summarized through a data integration circuit, and the effective case information is cached into a sparse signal output buffer; the data output interface circuit receives the information in the sparse signal output buffer, judges whether effective information needing to be output exists or not, and reads formatted data if the effective information exists. According to the invention, the output signal of the quantization circuit is discriminated so as to distinguish an effective case and a noise background, only pixel data containing an effective particle hit event is subjected to position correlation and coding, pixels not containing the effective hit event are skipped, and whether the data output interface outputs data to the outside is determined according to whether the effective case exists or not.

The invention provides a radar signal adaptive threshold generation method, which comprises the following steps of: when a threshold is generated, assigning a lower limit of a received system binding threshold as a current threshold value; opening a time window with the point number of M, and counting the time window with the point number of M: counting every M points as a counting period, and recording instantaneous amplitude values of a plurality of points in the time window in one counting period; if the instantaneous amplitude values of the majority of points in the time window do not exceed the current threshold, increasing the current threshold on the basis of the current threshold; if the instantaneous amplitude values of the majority of points in the time window exceed the current threshold and 1.25 times of the average level of the noise floor exceeds the current threshold, the current threshold is increased on the basis of the current threshold; and when 1.25 times of the average level of the noise floor is lower than the current threshold, the current threshold is lowered. The invention further provides a computer readable storage medium which is used for realizing the radar signal adaptive threshold generation method.

A magnetic shielding barrel noise floor evaluation method based on cross-spectrum analysis comprises the steps of data acquisition, simultaneous acquisition of noise data at the central position of a magnetic shielding barrel by using two independent atom magnetometers, windowing processing of discretized noise data of two channels by using a Hanning window, and reduction of near side lobe leakage. The method comprises the following steps: windowing two-channel data, performing Fourier transform on the windowed two-channel data to obtain frequency spectrum data, and finally performing frequency domain cross-correlation conjugate operation to calculate frequency domain characteristics of common-mode components of the two atom magnetometers and obtain a noise spectrum of the common-mode components of the two atom magnetometers, namely the noise floor of the magnetic shielding barrel. According to the method, common-mode components between the channels are extracted through cross-spectrum analysis operation, so that the noise floor of the magnetic shielding barrel or the noise floor of the magnetic shielding environment is evaluated, a metering type magnetometer with higher precision does not need to be researched, developed and used, the noise floor of the magnetic shielding barrel can be evaluated only through two magnetometers with lower sensitivity, the cost is lower, and the method is suitable for popularization and application. The equipment complexity of the extremely weak magnetic measurement device does not need to be increased, and the credibility is high.

A system that performs beam selection and beam merging using beam-specific signal quality metrics corresponding to a minimum noise floor. For example, a device may track a minimum noise floor for each beam, determine a highest minimum noise floor across the beams, and determine a noise floor ratio between the beam-specific minimum noise floor and the highest minimum noise floor. Using a combination of the noise floor ratio and signal-to-noise ratio (SNR) values, the device may perform beam selection by prioritizing low background noise as well as high SNR to select a pre-defined beam group. In addition, the device may use the noise floor ratio to perform beam merging and generate single-channel output audio data using the selected beam group. For example, the device may scale the beams based on a combination of the SNR value and the noise floor ratio.

The invention relates to the technical field of low-power-consumption electronics, and discloses a low-power-consumption instantaneous frequency measurement system which comprises a radio-frequency signal acquisition module, a signal time-frequency conversion module, a noise floor coding module and the like. The system generates a feature tensor through time-frequency analysis of a radio frequency signal, realizes dynamic threshold fusion in combination with noise floor sparse coding, extracts a multi-scale transient frequency feature by using a pyramid convolutional neural network, dynamically suppresses noise based on a signal-to-noise ratio weight, and reconstructs a real-time spectrum distribution image. The power consumption control module can adjust the ADC sampling rate and the clock frequency of the processing unit according to the signal strength, and the power management unit adopts multi-voltage domain independent power supply. The system solves the problems of frequency measurement precision and power consumption under complex noise, is suitable for the fields of wireless communication and the like, and has the characteristics of high precision, low power consumption, high spectral resolution, visualization and intuition.

A method of determining a location of a mobile device in the presence of a spoofing signal includes receiving a GNSS signal and detecting a spoofing condition based on a signal strength of the GNSS signal. The spoofing condition may comprise an increase in the signal strength, the signal strength exceeding a threshold value, an increase in a noise floor related to a measurement of the signal strength, or a combination thereof. The method may further comprise, responsive to detecting the spoofing condition, providing an indication of the GNSS signal as a spoofing signal.

The invention relates to the technical field of optical fiber sensing signal processing, in particular to a DAS signal floor noise elimination method, system and equipment based on statistical analysis, and the method comprises the steps: initializing system parameters, and loading a DAS application field data matrix; performing slicing and frequency domain transformation on the DAS original data; dynamically determining a peak threshold value, and performing adaptive peak searching according to the peak threshold value; sorting the identified and screened local maximum values in a descending order, and selecting frequencies corresponding to the local maximum values as main frequency candidates; performing clustering analysis on the selected dominant frequency candidates; calculating the cumulative proportion of each main frequency in the total statistics, selecting a center frequency, and forming the multi-frequency band elimination of the monitoring unit by using band elimination frequency bands formed by different center frequencies; and adding the multi-frequency band elimination to a statistical filtering frequency band set of all the point locations, storing the statistical filtering frequency band set of all the point locations as a statistical model, and performing background noise elimination on the DAS original signal based on the statistical model.

The invention discloses a vibration sensor background noise testing method based on a vacuum cavity, and the method comprises the steps: constructing an active-passive composite vibration isolation platform in a sealed vacuum cavity, rigidly fixing a to-be-tested sensor on the vibration isolation platform in a testing process, and guaranteeing that a sensitive axis of the to-be-tested sensor is orthogonal to a cavity reference; and an output signal is recorded for a long time at a sampling rate greater than or equal to 256Hz through a high-precision acquisition system. The data processing comprises polynomial detrending, sliding window de-averaging, multi-band elimination filtering and power spectral density (PSD) analysis so as to extract the equivalent amplitude and frequency domain characteristics of the background noise. In order to ensure the credibility of the result, the test needs to be repeated for three times, the consistency of the PSD curve is compared, and the deviation does not exceed 5%. The method combines the vacuum environment, composite vibration isolation, signal preprocessing and spectrum analysis technologies, can accurately quantify the background noise characteristics of the sensor, and is suitable for research and development verification and performance evaluation of the high-sensitivity sensor.

The invention discloses an optimal linear point searching method for a silicon-based Mach-Zehnder modulator. According to the method, the linearity of a corresponding working point is estimated according to the power of a fundamental component and a third-order intermodulation component of an output signal of the modulator, and the optimal linear point of the modulator can be obtained based on the maximum estimated value of the linearity. According to the method, a maximum entropy spectral subtraction method is adopted to reduce a signal noise floor, and a violent search method is adopted to retrieve an optimal combination of a modulator upper and lower arm phase difference and a reverse bias voltage to obtain an optimal working point. Compared with a conventional linearity test method that the linearity is obtained by testing when the working point of the modulator is set to be an orthogonal point, the optimal linear point searched by the method can obviously improve the linearity of the modulator, and a high-linearity silicon-based integrated microwave photon system can be realized.

The invention provides an electromagnetic shielding room special for an aviation product EMC test, and relates to the technical field of electromagnetic shielding rooms, the electromagnetic shielding room comprises a shielding shell, a manual shielding door and a ventilation waveguide window, the manual shielding door is arranged on the shielding shell for debugging personnel to go in and out, and the ventilation waveguide window is arranged on the shielding shell. The ventilation waveguide windows are installed on the side wall face and the top of the shielding shell respectively to achieve ventilation and electromagnetic shielding, the shielding shell is further provided with an overhaul shielding door and a filter, and the overhaul shielding door is used for connection and overhaul of cables in the shielding shell. During use, high-quality 2mm double-sided hot-dip galvanized steel plates are adopted for splicing, high-quality conductive gaskets are used for connecting splicing positions, so that good conductivity is guaranteed, high shielding performance is achieved, meanwhile, a filter is arranged for connecting cables entering and exiting the shielding shell, high-frequency signal components in a circuit are filtered out, a ventilation waveguide window with the cut-off frequency being 18 GHz is configured, and the shielding performance is improved. And the test requirements of aviation products with the working frequency range of 10kHz-18GHz floor noise are met.

The invention discloses a network signal interference detection system based on spectrum analysis. The system comprises a network signal acquisition module, a spectrum preprocessing module, a spectrum calculation module, a frequency separation spectrum kurtosis analysis module, a Bayesian online change point detection module, a parameter linkage module, an interference discrimination module and a result output module. The frequency separation spectrum kurtosis analysis module outputs kurtosis statistics and noise bottom estimation according to time slices based on the power spectrum; the Bayesian online change point detection module updates and outputs the change point position, the confidence coefficient and the adaptive time window parameter online; the parameter linkage module updates the frequency separation spectrum kurtosis analysis module according to the data; and the interference judgment module generates an interference event list, and the result output module outputs alarm and troubleshooting positioning data. According to the invention, automatic detection and discrimination of same-frequency superposition, adjacent-frequency superposition, burst intermittence and noise bottom uplift interference are realized, the accuracy is improved, and the troubleshooting cost is reduced.

The present invention discloses a base station construction plan guidance method, device and storage medium, the method comprising: determining a first test result that meets a first requirement by adjusting the isolation distance between a first node and a second node; the first requirement comprising: a noise floor rise value of the second node being a first target value; the first test result comprising: a first isolation distance between the first node and the second node; determining a second test result that meets a second requirement by adjusting the transmission power of the first node; the second requirement comprising: a noise floor rise value of the second node being a second target value; the second test result comprising: the transmission power of the first node; and determining a target isolation distance and / or a target noise floor rise value under a target rate requirement based on the first test result and the second test result.

The present invention provides a Halow communication module, and a mobile communication device and system. The Halow communication module comprises a Halow basic module, a radio frequency switch module, a radio frequency antenna module, and a Halow signal filter module. The radio frequency switch module is used for switching a radio frequency transmitting path or a radio frequency receiving path. The Halow signal filter module has a plurality of passbands which correspond to a plurality of frequency bands to which a Halow signal belongs. The mobile communication device comprises the Halow communication module, an LTE communication module, a WiFi communication module, and a communication control module. A Halow antenna is provided with a radio frequency filter which is used for inhibiting the noise floor of a transmitter of the Halow communication module. On the basis of a comparison result of communication network parameters of the communication modules and the corresponding thresholds, and the configured communication modules, the communication control module selects the priority and the communication mode of the mobile communication device. According to the present invention, the stable, reliable and high-quality communication process of the mobile communication device can be maintained in various network coverage scenarios, and stable and effective data transmission is achieved.

The present application discloses a method and system for reconstructing a reflected echo, which belongs to the field of signal processing technology. In the method for reconstructing a reflected echo of the present application, a complete reflected echo is composed of M frame, <h2 style=";text-align:left;direction:ltr">M×N sampling points, where each frame is sampled with a delay T Complete data sampling, sampling at equal intervals per frame N point, at each <h2 style=";text-align:left;direction:ltr"> A / D The interval time of the conversion cycle is used to realize synchronous data accumulation. When a complete reflection echo sampling is completed, the synchronous accumulation is also completed. M Frame sampling data reorganization, <h2 style=";text-align:left;direction:ltr"> Qavg The complete reflected echo is accumulated and averaged, logarithmically transformed, and curve fitted to achieve the reproduction of reflected echoes. Through data analysis, transmission loss and fault point detection are completed. This application meets the sampling accuracy conditions. <h2 style=";text-align:left;direction:ltr"> A / D Converter to 1 / M The sampling rate meets the performance requirements of the full load sampling rate, which can greatly reduce the <h2 style=";text-align:left;direction:ltr"> A / D The high-speed sampling rate requirements of the converter reduce the background noise, and the power consumption and cost of the signal analysis system are also greatly reduced.

The invention discloses a broadband random noise source with a high excess noise ratio, and belongs to the technical field of microwave and millimeter wave vacuum devices. The broadband random noise source comprises a medium-loaded gyrotron traveling wave tube and a Bragg filter, a wide-spectrum noise floor signal is generated by increasing the axial length of a medium loading circuit in the medium loading gyrotron traveling wave tube, and reflection filtering is performed on the wide-spectrum noise floor signal by arranging a Bragg filter between a collecting stage and an output window, so that a high-flatness wide-spectrum random noise signal is obtained. The characteristics of high power, high frequency and wide bandwidth of the gyrotron traveling wave tube for amplifying microwaves are utilized, noise floor signals can be amplified to generate high-power noise signals, meanwhile, the Bragg filter is loaded to carry out filtering and power balancing on the noise floor signals, and therefore high-power, wide-frequency-spectrum and high-flatness random noise signal output is achieved. And the output power of the generated noise signal is effectively improved.

This application discloses a wind tunnel-based noise detection method, apparatus, storage medium, and electronic device. The method includes: when testing a test piece in a wind tunnel, acquiring a measured noise spectrum in the testing environment and obtaining the operating frequency of the wind tunnel's inverter; matching the background noise spectrum corresponding to the operating frequency in a background noise spectrum library based on the operating frequency; aligning and dividing the measured noise spectrum and the background noise spectrum using a frequency band division rule adapted to the inverter's operating frequency to obtain multiple frequency band units; based on the acoustic energy density of the measured noise spectrum and the background noise spectrum within each frequency band unit, extracting the energy component corresponding to the background noise spectrum from the measured noise spectrum through directional orthogonal projection, and reconstructing the spectrum of the residual energy after extraction to obtain the effective aerodynamic noise spectrum of the test piece; and calculating the noise detection result of the test piece based on the effective aerodynamic noise spectrum.

A wireless device is configured to transmit radar signals at extremely low spectral density so that the receiver can be operated at the same time without saturating and thereby enabling full-duplex operation. To achieve sufficient range, beamforming is used together with a very long correlation time for detection of the radar signal, in which case the receiver can resolve backscattered or reflected radar signals far below the noise floor, while the bandwidth of the transmitted radar signal can be very high. The wide bandwidth enables the wireless device to maintain a low power spectral density while increasing output power, and also increase the depth resolution. The extremely low power spectral density even in the transmission beam direction makes the probability that communication will be disturbed very low. The high bandwidth calls for digital filters for both transmitted and received signals to equalize the radar frequency response characteristics of the radio unit.

Systems and methods for a MEMS microphone package are disclosed. The MEMS microphone package may include a first port to direct sound to a MEMS system, including a die substrate, an acoustic membrane, and one or more plates. The MEMS microphone package may include an ASIC to produce microphone output based on an electrical signal, a PCB, a lid, and a second port. The first port and the second port may define a front volume and a back volume respectively. The second port may increase the back volume to improve sensitivity and reduce the acoustic sensor's noise floor to improve signal-to-noise ratio.

[Problem] To provide a determination device, a determination method, a determination program, and a radar device for determining the presence or absence of radio wave interference that does not cause a noise floor rise even when there is radio wave interference and does not require a special pattern provided for radio wave interference. [Solution] A determination device according to one embodiment of the present technology is provided with a reception unit and a determination processing unit. The receiving unit receives a first reflected chirp signal generated by a first transmission chirp signal transmitted toward an object by a transmitting device and reflected from the object, and a second reflected chirp signal generated by a second transmission chirp signal transmitted by the transmitting device at a time different from a time at which the first transmission chirp signal is transmitted and reflected from the object. A determination processing unit compares a plurality of arrival angle spectrums generated based on a plurality of corrected signals obtained by performing a plurality of velocity aliasing correction processes on the first reflected chirp signal and the second reflected chirp signal, and determines whether or not a useless signal is received.

This invention discloses a magnetotelluric denoising method and system based on a physics-driven and conditional diffusion model, belonging to the field of electromagnetic signal processing technology. This method integrates the advantages of neural networks and the physical characteristics of magnetotelluric signals, and proposes a physics-driven synthesis strategy based on statistical priors of measured source data and a curriculum-based dynamic physics-driven noise addition strategy based on training progress factors. Addressing the scarcity of high-quality, clean samples, a clean background field that highly approximates real physical characteristics is synthesized using a parametric mathematical model. The introduction of curriculum-based dynamic physics-driven noise addition (such as noise floor annealing, square wave incrementing, and morphological clustering pulses) accurately simulates natural background degradation and complex human interference, enabling the model to adapt to extremely complex time-frequency characteristics under different geological structural backgrounds. This completely overcomes the domain shift problem caused by static Gaussian white noise training, constructs a richer sample library, and lays the foundation for accurate denoising of conditional diffusion models.